Assessing social responsibility: A quantitative analysis of Appraisal in BP’s and IKEA’s social reports

Abstract

A growing public awareness of the potential negative impacts of corporate activities on the natural environment and society compels large companies to invest increasing resources in the communication of their responsible conduct. This article employs Appraisal theory in a comparative analysis of BP’s and IKEA’s 2009 social reports, each company’s record of their non-financial performance. The main objective is to explore how, through Appraisal resources, BP and IKEA construct their corporate identity and relationship with their stakeholders. The analysis reveals two markedly different approaches to the construction of a responsible corporate identity. While BP deploys interpersonal resources to portray itself as a trustworthy and authoritative expert, IKEA discloses itself as a sensitive and caring corporation, engaged in a continual effort to improve. These differences are interpreted in light of the legitimization challenges the two companies face.

Keywords

Appraisal corporate identity corporate social reports corporate social responsibility corpus-based approaches to evaluation discourse analysis evaluation legitimacy theory stance

1. Introduction

High-profile industrial accidents, environmental disasters, corporate accounting and human rights scandals have contributed to exposing the tensions that often exist between profit-oriented activities and broader social values. As Deegan et al. (2000: 101) note, ‘business today operates in a climate of intense public scrutiny’, where stakeholders expect ‘that corporations and industries accept accountability for the social and environmental implications of their operations’.

Organizations have responded to ‘society’s heightened sensitivity to the externalities of business activities’ (Ullmann, 1985: 540), releasing a growing amount of information concerning the social and environmental impacts of their operations and documenting their efforts to implement social responsibility principles. New forms of corporate ‘social’ reporting have emerged and have rapidly become standard practice among larger companies.

While the phenomenon of corporate social reporting has been extensively investigated by accounting and management scholars, discourse analysis research in this area is still limited. Moreover, previous discourse analyses of corporate social reports have primarily focused on their ideational function (Fairclough, 1992), namely how these texts construe a particular representation of reality. Despite the widely acknowledged role of social reports as vehicles for the interaction and ‘dialogue’ between organizations and society (Gray et al., 1995), their interpersonal functions (Fairclough, 1992), that is, how through these texts reporting companies construct their identity and relationship with their interlocutors, have not been the object of systematic investigation.

This article aims to partially fill this gap by employing Appraisal theory in a comparative analysis of interpersonal language resources in BP’s and IKEAS’s 2009 social reports. The analysis aims to address the following questions: a) How do BP and IKEA discursively construct their corporate identity? b) What kind of relationship do they establish with their putative interlocutors? In order to arrive at a fuller understanding of the potential communicative motives and consequences for the use of Appraisal resources in these companies’ reports, the results are interpreted through the lenses of legitimacy theory, the current dominant explanatory framework for social reporting, which identifies the need for legitimating an organization’s actions as the primary driver behind the disclosure of social and environmental information.

The article aims to provide a more systematic consideration of the interpersonal functions of social reports, if limited to the case of BP’s and IKEA’s reports. More broadly, by combining legitimacy theory and Appraisal analysis, this article attempts to shed some light on the discourse mechanisms underlying the ongoing renegotiation of roles and relations at the interface between business and society, in the wake of the current intense public debate over the social responsibilities of business (Carroll and Shabana, 2010).

The article is structured as follows. The section immediately below provides a brief overview of legitimacy theory. Section 3 reviews previous relevant discourse analyses of social reports, section 4 describes the Appraisal theoretical framework and section 5 outlines the methodology employed for the analysis. The last two sections present the results and provide some interpretative conclusions respectively.

2. Legitimacy theory

Voluntary social disclosures by large companies have constantly grown in volume over the last two decades (Bebbington et al., 2008; Deegan, 2002; Gray et al., 1995, 1996, 2001; Milne et al., 2009). Social and environmental information included in annual financial reports is now frequently accompanied by stand-alone social reports (Bebbington et al., 2008; KPMG, 2008), which are generally published on the companies’ websites. The BP and IKEA reports analysed in this article are examples of this kind of disclosure.

Corporate social disclosures vary greatly in content and form (Gray, 2002; Gray et al., 1995, 1996). As Gray et al. (1996: 82) contend, differently from financial reporting, limited institutional regulation makes social reporting ‘a virtually limitless area of potential activity’. Topics covered usually encompass the natural environment, employees, the local community and consumer issues, and information is normally presented in a combination of narrative statements, quantitative data and financial information (Gray et al., 1996).

The evolving phenomenon of voluntary corporate social reporting has attracted considerable interest among accounting researchers, especially since the mid-1990s (Deegan, 2002). Attempts to explain and theorize this phenomenon have yielded a vast and heterogeneous literature (Gray et al., 1995; Thomson, 2007). Within social and political theory studies (Gray et al., 1995), social disclosures have been interpreted from political economy (e.g. Adams and Harte, 1998; Arnold, 1990; Tinker et al., 1991), stakeholder (e.g. Roberts, 1992; Ullmann, 1985) and legitimacy perspectives. While there is still no single universally accepted theory of social reporting, legitimacy theory has emerged as the preeminent framework for explaining organizations’ social disclosure practices (Bebbington et al., 2008; Campbell et al., 2003; Deegan, 2002; Gray et al., 1995).

According to legitimacy theory, social reporting is primarily motivated by the need to justify an organization’s activities and obtain social legitimization. The theory assumes the existence of a social contract (Shocker and Sethi, 1973) governing the relationship between business and society. In order to be considered legitimate and thus continue to exist and grow, organizations are expected to behave within the bounds of what is deemed acceptable and desirable within the broader social system (Dowling and Pfeffer, 1975; Shocker and Sethi, 1973). A breach of the social contract may have a number of negative consequences for a business, including reduced demand for its products by consumers, restricted access to labour supply and financial capital, increased pressures from lobbies and more stringent government regulation and control (Brown and Deegan, 1998; Deegan et al., 2000). Failure to comply with the terms of the social contract may threaten the organization’s legitimacy and ultimately lead to the withdrawal of its licence to operate (Tilling and Tilt, 2010).

Given that social expectations change over time, an organization will need to be responsive and take action in order to maintain its legitimacy (Suchman, 1995). Communication, and thus disclosure, plays a fundamental role in this process. As social expectations change, the organization must provide evidence that it is changing accordingly or, as Deegan et al. (2000: 105) explain, ‘the community may, as a result of incomplete information, still initiate actions which will not be in the interests of the organization’. Companies thus react to society’s changing expectations by disclosing information directed at (re-)establishing congruence between social norms and values and organizational conduct, thus ‘narrow[ing] the legitimacy “gap” between how the organization wishes to be perceived and how it actually is perceived’ (Campbell et al., 2003: 561).

Several studies have provided empirical support for legitimacy theory. Indeed, a significant positive correlation has been found between companies’ increased voluntary social disclosure and the following factors: industrial accidents (Deegan et al., 2000; Patten, 1992); negative media exposure (Brown and Deegan, 1998; Deegan et al., 2002); changing environmental legislation (Buhr, 1998); environmental lobby groups’ growing concern (Deegan and Gordon, 1996); and successful environmental prosecution (Deegan and Rankin, 1996). Additional support is provided by O’Donovan (2002) who, employing semi-structured interviews, found management decision-making to be compatible with a legitimacy explanation.

The perspective of social reporting provided by legitimacy theory is particularly suited to the aims of this study. Given that ‘legitimacy represents a relationship with an audience’ (Suchman, 1995: 594, emphasis in original), and assuming the need for establishing such a relationship as an important determinant of the companies’ reporting behaviour, legitimacy theory is likely to offer valuable insights into the communicative implications and motives for the use of interpersonal language resources in BP’s and IKEA’s reports. However, studies adopting legitimacy theory have tended to focus on the information content of social disclosures, analysing them by using content analysis (Krippendorff, 2004). The role of language in the construction of a positive corporate identity and a favourable, legitimating company–stakeholder relationship is still under-researched. Indeed, as Fairclough (1992, 2003) suggests, language plays a crucial role in the texturing of social identities and interpersonal relations. By focusing solely on what kind of information is exchanged, only a partial picture of the communicative function performed by social reports can be obtained. Through the analysis of interpersonal language resources in BP’s and IKEA’s reports, this article aims to offer an alternative perspective on the process of organizational legitimization which, according to legitimacy theory, social reports facilitate.

3. Previous discourse analyses of social reports

Several studies have analysed social reports from a social constructionist perspective, based on the understanding of discourse as ‘a practice not just of representing the world, but of signifying the world, constituting and constructing the world in meaning’ (Fairclough, 1992: 64). Livesey (2002: 321) analyses ‘salient themes, metaphors, modes of expression, and argument structures’ employed by Shell in its first stand-alone social report. The author explores how, in the wake of mounting public pressure after major environmental controversies, the company reshaped its public discourse. Livesey (2002: 338) concludes that Shell’s discourse reflected a ‘desire for hegemonic control’ over the rules for the resolution of such controversies, adopting a conception of sustainable development which failed to pose any significant challenge to the ‘taken-for-granted truths about the nature of markets, competition, and economic actors’ (Livesey, 2002: 331).

Laine (2005) employs discourse analysis to deconstruct the meaning of the term ‘sustainable development’ in the social disclosures produced by Finnish listed companies. The results of his analysis, in line with Livesey (2002), show how sustainability is constructed as compatible with economic growth and characterized as a ‘win–win’ solution to social and environmental problems that requires no major restructuring of the current economic paradigm. This finding is confirmed by Laine (2010), who adds a diachronic perspective to the analysis of the business discourse of sustainability. Through a longitudinal study of social disclosures published by three major Finnish companies over a period of almost 20 years, the author observes a gradual shift from a view of sustainability as a strong and problematic phenomenon to a feasible objective, coherent with current business practices.

As Laine (2005) notes, sustainability, however, is often reduced in social reports to abstract and vague principles, with relatively little emphasis on the concrete measures adopted to achieve it. A similar conclusion is reached by Milne et al. (2005). In their account of the employment of the journey metaphor in the social reports of leading New Zealand organizations, the authors note how, by conceptualizing sustainable development as a journey, the reporting companies avoid facing the complex and highly problematic issue of identifying and describing sustainability goals and actions to achieve them. In a recent study, Lischinsky (2010) employs corpus-based techniques to explore the business notion of sustainable development in a collection of reports issued by 50 large Swedish companies. By analysing the collocational and prosodic profile of the key lemma sustain*, the author concludes that the business view of sustainability is frequently connected with profit and corporate prestige, and that market objectives and proclaimed intentions feature more prominently than reference to concrete actions.

Fairclough (1992: 8) argues that ‘[c]hanging discourse practices contribute to change in knowledge [. . .], social relations and social identities’. Drawing on the systemic functional tradition (Halliday, 1978, 1994), the author thus advocates a multifunctional approach to discourse analysis that accounts for how texts simultaneously represent reality – the ideational function, enact social relations – the relational function, and establish identities – the identity function. The studies briefly reviewed above mainly concentrate on the ideational function of social reports, that is, how these texts constitute and naturalize a certain representation of reality. Specific research into the interpersonal (identity and relational) functions of social reports is still relatively sparse, though aspects connected with these functions have been considered in several studies. In their comparative analysis of the social reports published by The Body Shop International and by the Royal Dutch/Shell Group, Livesey and Kearins (2002: 246) note that ‘the report’s language was replete with descriptions of personal feeling such as “caring,” “wanting,” “striving,” “determination,” and “pride”’, and that ‘metaphors of the heart were employed to describe the corporation’ and construct a certain ‘corporate persona’. Buhr and Reiter (2006) discuss the combination and mingling of environmental philosophies in the social reports released by the mining company Noranda, in relation to its aim to ‘generate a green identity of the company and legitimate the company with various publics’ (Buhr and Reiter, 2006: 10). In her genre analysis of a corpus of social reports, Skulstad (2008) describes rhetorical strategies employed in these texts to serve the communicative aim of creating and projecting a positive, ‘green’ image of a company and contrasting unfavourable ‘externally-constructed’ images. These studies, however, mostly focus on broad rhetorical patterns, narrative story lines and themes employed in social reports and do not engage in an explicit and systematic analysis of the language resources connected with the interpersonal functions of these texts. With reference to the framework for discourse analysis proposed by Fairclough (1992), they do not provide an account of the interpersonal functions of social reports at the ‘discourse as text’ level, that is, at the level of language forms and meanings.

Fairclough (1992) proposes investigating the interpersonal functions of discourse by analysing interactional control (e.g. turn-taking systems in conversation), modality, politeness and ethos. In Fairclough (2003: 171), the author extends the range of resources considered to include evaluation, that is, ‘the explicit or implicit ways in which authors commit themselves to values’. Appraisal theory (Martin, 2000; Martin and White, 2005), developed within the systemic functional tradition, provides a comprehensive framework for the systematic investigation of those textual features which realize the interpersonal functions of discourse, which combines the resources for the expression of feelings, attitudes, evaluations and modality into a unified model.

4. The Appraisal framework

Appraisal (Martin, 2000; Martin and White, 2005) is a comprehensive term encompassing all the linguistic resources for the expression of affect, judgements, assessments and for the negotiation of stances and ideological positions (White, 2001). Appraisal theory¹ originated as an extension of the theoretical linguistic framework of Systemic Functional Linguistics (Halliday, 1994; see also Eggins, 2004). Appraisal groups and categorizes interpersonal meanings (Halliday, 1994), the resources which language users employ to manage interpersonal roles and relationships, to establish empathy and negotiate solidarity and alignment.

The Appraisal framework organizes interpersonal meanings into different systems (Figure 1). The most important distinction is between:

Attitude, which involves the linguistic expression of emotional dispositions, ethical judgements and evaluations according to aesthetic and significance criteria;

Engagement, which comprises the resources by which language users negotiate the arguability of their utterances; and

Graduation, which groups the linguistic means that are used to modulate the intensity of assessments.

Figure 1.

An overview of Appraisal (adapted from Martin and White, 2005)

Each subsystem is further divided into different subcategories. As Hood (2004: 74) summarizes, a principle of delicacy applies to the different subsystems of Appraisal as to any system network within Systemic Functional Linguistics. The movement from left to right across the networks indicates the passage from general to more specific analysis.

The study presented in this article focuses on the systems of Attitude and Engagement. For this reason, the next sections concentrate on these categories without providing a detailed account of Graduation.

4.1. Attitude

Attitude resources are subdivided into three subcategories:

Affect, which concerns the expression of feelings, emotions and states of mind;

Judgement, which comprises positive and negative normative assessments of human behaviour;

Appreciation, which involves the positive and negative evaluation of natural or semiotic phenomena, entities and processes.

Given the semantic nature of this categorization, the instantiation of the sub-classes can involve a wide range of lexico-grammatical resources. Affect may be expressed through verbs of emotion (to love/to hate, to frighten/to reassure), adverbs (happily/sadly), adjectives (angry/pleased) or nominalized forms (joy/despair, confidence/insecurity) (White, 2001). Judgement, as noted by Martin (1995), tends to be realized adjectivally (unjust/fair-minded, skilled/incompetent), although it can also be conveyed through nominalized forms (best behaviour, outstanding performance) and prepositional phrases (in order, in shape). Similar considerations apply to the category of Appreciation (extremely beautiful/breathtaking beauty, challenging/challenge). The following short excerpts exemplify the different types of Attitude in the social reports analysed.

Affect (emotions, feelings, states of mind)

(1) IKEA is concerned about climate change and works to minimize carbon dioxide emissions from all parts of our operations.

(2) However, we still have work to do and I deeply regret the two deaths we suffered in operations at BP sites last year.

Judgement (capacity, tenacity, ethical commitment)

(3) BP is progressive, responsible, innovative and performance driven.

(4) We believe in doing more with less, being humble by listening to and learning from others and being fair and honest.

Appreciation (aesthetics, value, relevance)

(5) Our business idea is to offer a wide range of well designed, functional home furnishing products.

(6) We invest in biofuels research because we believe there is scope for many more advanced, sustainable products.

4.1.1. Inscribed versus evoked realizations

An important dichotomy in Appraisal theory is between inscribed and evoked instances. The former refers to the cases in which an evaluation is explicitly signalled through the use of manifestly positive or negative words or expressions. As (White, 2001: 4) explains, ‘under explicit Attitude we can point to overtly evaluative/attitudinal words or combinations of words [. . .] which unproblematically carry a positive or negative sense’. Evoked Attitude is conveyed instead by ideational meanings, factual information that has the capacity, in a given culture or context, to evoke a judgemental response (White, 2001). Compare, for example:

(7) BP is driving innovative, efficient and responsible operations.

(8) In FY09, IKEA Sweden and WWF Sweden organised a sustainability contest for children in grades 1–3 to teach them about natural resources and recycling.

In example 7, the adjectives innovative, efficient and responsible explicitly convey the writer’s positive judgement on BP’s operations. In example 8, a positive judgement on IKEA’s attitude towards the environment might be inferred based on the assumption that teaching children to recycle and to take care of natural resources is a good thing to do. However, no explicit evaluation of IKEA’s behaviour is proffered (e.g. ‘IKEA is responsible towards the environment and attentive towards children’s education’).

However, as Bednarek (2006: 31) highlights, the distinction between inscribed and evoked Appraisal ‘is far from clear-cut’ and should be better conceived of as a cline. In fact, the level of explicitness of the positive or negative evaluative connotation of lexical items is often difficult to determine conclusively, as it depends upon co-textual factors – that is, the particular position of a lexical item in the text, contextual factors – for example, the text type and the communicative situation, and on the receiver’s ideological ‘reading position’ (Martin, 1995; Martin and White, 2005) – that is, which individual values and criteria are brought into play in the interpretation of a text (see also Eggins and Slade, 2005).

The analysis of Appraisal is further complicated by the fact that inscriptions can, in certain cases, evoke secondary interpretations at different levels (Page, 2003). Consider, for example:

(9) We are also developing plans for pioneering power plants with carbon capture and storage.

In this example, the positive Appreciation of the target power plants might simultaneously evoke an indirect positive Judgement on BP’s technical capabilities, based on the assumption that if BP is able to develop advanced plants it must possess very high technical skills.

The theorized possibility for attitudinal stances to be evoked either a) by superficially neutral, ideational meanings or b) as secondary codings deriving from explicit lexicalizations, adds complexity to the analysis and dramatically increases the degree of subjectivity involved in it, posing important challenges to a reliable quantification of Appraisal items. The measures adopted in this article to deal with these challenges are described in section 5.

4.2. Engagement

The system of Engagement groups together a wide array of lexico-grammatical resources that have been studied separately under such headings as evidentiality (Chafe and Nichols, 1986), hedging (Hyland, 1996) and modality (Hoye, 1997; Palmer, 2001). Within the Appraisal framework, these resources are associated with a single, complex discourse function, that of expressing the writer’s intersubjective stance with respect to the various opinions and value positions referenced in a text (White, 2003).

The interpretation of the functioning of Engagement in discourse is centred around the notion of dialogism. The theory assumes Bakhtin’s ‘dialogical principle’ (Bakhtin, 1981), according to which every utterance, even in monologic written texts, references or anticipates the responses of actual or imagined readers. Engagement resources construe the writer’s position with respect to the ‘voices’ echoed in text, by entertaining or refusing putative counterarguments and alternative opinions.

Appraisal theory distinguishes between monoglossic and heteroglossic utterances (Figure 2). The former are bare assertions, such as:

(10) SPARSAM low energy bulbs are a great alternative to traditional incandescent bulbs

Figure 2.

Engagement: monogloss versus heterogloss (adapted from Martin and White, 2005)

where no alternative viewpoint is referenced. Heteroglossic utterances, on the other hand, are those in which the writer’s engagement with alternative positions is signalled by the use of epistemic or modal markers, as in the example below:

(11) IKEA co-workers are often on-site at these suppliers, and we believe our close and long-term relationships with them have contributed to many improvements.

Martin and White (2005: 99) stress that, although monoglossic assertions do not explicitly mark the writer’s engagement with alternative points of view, they maintain their dialogic functionality. Through an utterance of this kind, the writer ‘presents the current proposition as one which has no dialogistic alternatives that need to be recognized in the current communicative context, [. . .] and hence capable of being declared categorically’ (Martin and White, 2005: 99).

The resources for heteroglossic Engagement are broadly subdivided into those that act to challenge or refute alternative viewpoints, dialogic contraction, and those that open the dialogic space of the text to competing perspectives, dialogic expansion. This difference is illustrated in the following examples:

(12) Audits have proven good progress in Pakistan and Bangladesh.

(13) We’ve reviewed more than 20 technologies, some of which may well become significant in the future.

Although both utterances are heteroglossic, example 12 fences off potential disagreeing arguments by means of the verb prove, whereas the modal may in example 13 frames the proposition as but one hypothesis and thus exposed to dialogic confrontation.

Engagement is further subdivided into four main categories (Figure 3). Resources for dialogic contraction are classified under the headings of Disclaim and Proclaim. The former includes negation and expressions of counter-expectation, while the latter comprises epistemic markers of certainty and expressions of endorsement of attributed propositions.

The following are examples of these categories:

Figure 3.

Engagement: dialogic contraction and expansion (adapted from Martin and White, 2005)

Disclaim

(14) IKEA does not accept corruption in any form, whether direct or indirect, and works proactively to prevent it.

Proclaim

(15) There’s no doubt that our assets are working together more effectively thanks to OMS.

The resources for dialogic expansion are subdivided into the categories of Entertain, which includes epistemic markers of doubt and modals of possibility, and Attribute, wordings for the neutral acknowledgement of or the distancing from reported propositions. These categories are exemplified in the following excerpts from the reports analysed:

Entertain

(16) We are also enhancing our capabilities in natural gas, which is likely to be a vital source of relatively clean energy during the transition to a lower-carbon economy and beyond.

Attribute

(17) In its 2009 report, the panel said Tangguh had brought tangible benefits to the area.

5. Methodology

This study presents a comparative analysis of Appraisal in two long texts, BP’s ‘Sustainability Review 2009’ and IKEA’s ‘Sustainability Report 2009’.² Their main features are summarized in Table 1.

Table 1.

Corpus details

	Tokens	Types	Pages
BP	24,300	3741	37
IKEA	30,289	3365	74

The analysis is primarily quantitative and focuses on the Appraisal systems of Attitude and Engagement. As far as Attitude is concerned, the analysis is confined to explicit realizations. Evoked instances were not considered in the analysis. In view of the primarily quantitative aims of this study, it was judged that the costs of a higher degree of subjectivity in the analysis process and of a more difficult computability of the results greatly outweighed the benefits in terms of detail and comprehensiveness that could have been achieved by including them in the analysis. In the case of Engagement, the analysis is restricted to the subcategories of Proclaim and Entertain.

Two distinct methodologies were employed for the analysis of Attitude and Engagement. The analysis of Attitude is based on the computer-aided manual annotation of the corpus.³ The annotation was performed on the raw, segmented (‘tokenized’) texts imported into a spreadsheet document. The texts were automatically annotated with Part of Speech labels⁴ to facilitate the consolidation of the results. Instances that were judged by the analyst to explicitly and unproblematically carry a positive or negative Appraisal value were marked and classified according to the guidelines and examples provided in Martin (2000) and Martin and White (2005) and drawing on coding choices outlined in other studies that apply the Appraisal framework. When the coding was complete, annotated items were automatically retrieved through filters for the production of statistics and the creation of frequency lists.

In order to control for bias and subjectivity in the identification of instances (Hunston, 2004), inter-coder agreement (Artstein and Poesio, 2008) was measured on a sample excerpt from the corpus.⁵ The reliability test was carried out in two distinct phases and involved two independent non-expert annotators other than the author. Each test was preceded by a training session during which the criteria for the identification of Appraisal items were explained to the reliability coder and discussed with her thoroughly. The results from the first test highlighted areas for improvement in the guidelines provided to the annotator. Most observed disagreement was related to the often blurred distinction between inscribed and evoked Attitude (see section 4.1.1). Words eliciting positive associations or describing positive facts as well as lexis indicating quantities, amounts or extents were often marked as evaluative.⁶ Two examples are reported below.

(18) In FY09, IKEA Sweden and WWF Sweden organized a sustainability contest for children in grades 13 to teach them about natural resources and recycling.

(19) During the year the filling rates increased and based on our work to increase the awareness along the value chain we expect a continued positive [agreed] development.

The second test was preceded by a longer and more accurate training session, based on improved annotation guidelines. Upon completion of the coding task, the annotation made by the external coder was compared with that made by the present writer independently. To calculate inter-coder agreement, weighted Kappa statistics were adopted (Cohen, 1960). Results are summarized in Table 2. Test 1 resulted in a chance-corrected coefficient of agreement of k = 0.51, interpreted as ‘moderate’ according to Landis and Koch’s (1977) scale. Test 2 yielded a value of k = 0.62, which indicates a ‘substantial’ level of agreement. As this second score is the result of a more careful and in-depth training, it can be considered a more trustworthy assessment of the level of inter-coder agreement that can be attained in the task considered.

Table 2.

Inter-coder agreement results

	Percentage sample size	Observed agreement	k coefficient	Assessment
Test 1	3.45	0.93	0.51	Moderate
Test 2	4.48	0.98	0.62	Substantial

For the analysis of Engagement two collections of potential markers were assembled adapting and integrating the lists of stance markers of certainty (adapted to Proclaim) and doubt (adapted to Entertain) provided in Biber and Finegan (1989). The two lists were used to filter the texts through an automatic procedure which detects and prints the frequency counts of the markers. The markers were subsequently manually checked in the corpus to disambiguate their meaning and eliminate from the count those that were not used as Engagement resources. The following is an example of this ambiguity:

(20) (a) As the UN’s Food and Agriculture Organization has shown, the main causes of deforestation are not biofuels.

(b) The energy quiz and facts show how to save energy at home, at work and at play.

5. Results

The results of the analysis show significant differences in the use of Appraisal in the two reports. Examination of the 25 most frequent instances of Attitude in the corpus⁷ (Table 3) reveals that the sum of the number of occurrences per thousand words of inscriptions which positively refer to technical competence and know-how (highlighted in bold) is substantially higher in BP’s report (3.10 against 0.80). This difference is reflected in the quantitative data on the distribution of the categories of Attitude (Table 4), which indicates that instances of positive Judgement are comparatively more frequent in BP’s text. Table 3 highlights, on the other hand, a comparatively stronger emphasis on sustainability, improvement⁸ and volition/intentionality in IKEA’s report (inscriptions in small caps).

Table 3.

Attitude: General wordlists

BP			IKEA
Inscr.^a	Fr.^b	Prop.^c	Inscr.^a	Fr.^b	Prop.^c
improvement	1.07	4.02	sustainable	2.48	9.12
key	1.03	3.86	improvement	2.11	7.79
challenge	0.95	3.55	good	1.91	7.06
skill	0.74	2.78	want	1.32	4.87
sustainable	0.74	2.78	challenge	1.02	3.77
good	0.70	2.63	safe	0.73	2.68
advanced	0.66	2.47	key	0.69	2.55
efficient	0.66	2.47	important	0.59	2.19
safe	0.58	2.16	positive	0.56	2.07
improved	0.53	2.01	improved	0.53	1.95
progress	0.53	2.01	progress	0.49	1.82
important	0.49	1.85	responsible	0.46	1.70
clear	0.45	1.70	specialist	0.46	1.70
commitment	0.45	1.70	complex	0.43	1.58
responsible	0.45	1.70	quality	0.43	1.58
right	0.45	1.70	strive	0.43	1.58
expertise	0.41	1.55	actively	0.40	1.46
significant	0.41	1.55	commitment	0.40	1.46
capability	0.34	1.24	efficient	0.40	1.46
concern	0.34	1.24	healthy	0.36	1.34
consistent	0.34	1.24	right	0.33	1.22
commit	0.29	1.08	well	0.33	1.22
effective	0.29	1.08	easy	0.30	1.09
priority	0.29	1.08	value	0.30	1.09
reliable	0.29	1.08	commit	0.26	0.97

Attitudinal inscription; ^boccurrences per thousand words; ^cpercentage proportion over total inscription count.

Table 4.

Subtypes of attitude: Distribution^a

	Affect		Judgement		Appreciation
	+	−	+	−	+	−
BP	2.63	2.32	54.80	1.55	32.51	6.19
IKEA	8.64	2.19	49.88	0.49	32.00	6.81

Proportion of total inscription count (%); χ² = 28.0, df = 5, p < 0.001.

These differences become strikingly evident if we filter the results to display only the most frequent instances of authorial Affect, the emotional dispositions expressed in the first person, and self-Judgement, the companies’ direct evaluation of their own qualities and performance. As shown in Table 5, BP makes persistent use of evaluative language to positively refer to its technical strengths and expertise. Among the most frequent instances, the wordlist features skill, expertise, capability, efficient, innovative and breakthrough (highlighted in bold). Consider, for example:

(21) Our frontier skills are being applied in new contexts such as our businesses in Azerbaijan and our exploration activity in Libya.

(22) Canadian oil sands. Leveraging our technical expertise to deliver a major energy resource responsibly.

Table 5.

Authorial Affect and self-judgement: Wordlists

BP			IKEA
Inscr.^a	Fr.^b	Prop.^c	Inscr.^a	Fr.^b	Prop.^c
improvement	0.49	6.06	improvement	1.35	17.15
commitment	0.45	5.56	want	1.16	14.64
skill	0.45	5.56	commitment	0.40	5.02
progress	0.37	4.55	strive	0.40	5.02
expertise	0.34	4.04	actively	0.36	4.60
capability	0.29	3.54	specialist	0.30	3.77
commit	0.25	3.03	value	0.30	3.77
improved	0.25	3.03	ambition	0.23	2.93
responsible	0.25	3.03	commit	0.20	2.51
safe	0.21	2.53	progress	0.20	2.51
actively	0.16	2.02	active	0.13	1.67
efficient	0.16	2.02	strong	0.13	1.67
innovative	0.16	2.02	expert	0.10	1.26
strive	0.16	2.02	forward	0.10	1.26
well	0.16	2.02	generous	0.10	1.26
consistent	0.12	1.52	hate	0.10	1.26
responsibly	0.12	1.52	know-how	0.10	1.26
right	0.12	1.52	sustainable	0.10	1.26
rigorous	0.12	1.52	systematically	0.10	1.26
value	0.12	1.52	concern	0.07	0.84
breakthrough	0.08	1.01	cost-consciousness	0.07	0.84
care	0.08	1.01	efficient	0.07	0.84
close	0.08	1.01	humbleness	0.07	0.84
confidence	0.08	1.01	improved	0.07	0.84
diversity	0.08	1.01	persistent	0.07	0.84

Attitudinal inscription; ^boccurrences per thousand words; ^cpercentage proportion over total inscription count.

The most frequent instances in IKEA’s report emphasize instead the company’s volition and tenacity. The modal want is the most frequent instance of authorial-Affect in IKEA’s text. This verb is extensively used to underline the company’s desire to achieve positive goals and is frequently followed by other positive evaluative items, as in:

(23) We want our co-workers to have a fulfilling career and grow with us.

(24) We also want to find ways to better utilize recycled materials in new products.

Other inscriptions, like strive, actively, commitment and ambition, underscore the company’s perseverance towards achieving desirable goals, as in:

(25) We strive to always use materials, surface treatments and production techniques with the lowest possible emissions.

(26) As part of our commitment to being a good neighbour, most IKEA stores are actively involved in their local communities.

The differences in the use of inscriptions in comparative and superlative form reinforce the contrast between BP’s emphasis on capacity and performance and IKEA’s focus on progress. Comparative and superlative morphology belong to the resources for the adjustment of the intensity or force of evaluations, which are grouped under the heading of Graduation. While superlatives clearly raise the intensity of appraisals, comparatives rather function as ‘hedges’. Compare, for instance, the following claims:

(27) (a) These give us a more accurate picture of suppliers’ factories.

(b)* These give us an accurate picture of suppliers’ factories.

When making assessments, BP uses superlatives more frequently, while IKEA employs comparatives relatively more often (Table 6). In BP’s report, superlatives are frequently used to highlight the company’s technological excellence, as in:

(28) We have three of the world’s most advanced CCS projects in our portfolio.

Table 6.

Attitude: Markers of comparison^a

	Comparatives	Superlatives
BP	3.86	4.17
IKEA	14.36	2.67

Proportion of total inscription count (%); χ² = 26.2, df = l, p < 0.001.

or to describe the operating environment as extremely demanding, thereby foregrounding BP’s technical reach, as in:

(29) We have decades of experience of using cutting-edge skills and technology to undertake complex oil and gas projects in many of the world’s most technically challenging and hostile environments.

Indeed, the very title of the report, ‘Operating at the energy frontiers’, accentuates the opposition between BP’s technical capabilities and the harsh environments where it operates by means of the ambivalent metaphorical term frontiers, which can be read as referring to both the most complex and formerly unreached operating areas, the frontiers of oil and gas extraction, and to the most advanced energy production technologies, the frontiers of technology.

Superlatives are also frequently employed to qualify BP’s opinion on issues of paramount interest to policy-makers and of broad societal relevance, first and foremost the future of energy supplies and global warming. Consider, for example:

(30) Opening up oil and gas fields to a range of potential competitors encourages the most efficient solutions.

(31) BP believes that a diverse energy mix including fossil fuels and renewables, produced and used efficiently, is best able to meet demand affordably at the same time as providing security of supply and addressing the issue of climate change.

What is noteworthy in these kinds of claims is not just their extreme relevance and potential implications, but also the way they are presented. Through the use of superlatives, even in the context of a dialogic expanding statement like 31, BP raises the ‘intensity’ of its evaluative statements. In doing so, the company makes a stronger claim to its epistemic right to make those assessments. In other words, by advancing stronger viewpoints on issues of general interest, BP implicitly asserts to be competent in them.

IKEA’s report, on the other hand, contains a higher number of evaluative items in comparative form. Comparatives relativize assessments. In the case of positive evaluations, the use of comparatives implicitly suggests the possibility of further betterment. For example, IKEA’s statement that:

(32) Supplier records now include water and energy use so that we get a better map over the Catalogue’s total environmental footprint

implies that the company’s control over the footprint of its catalogue has improved, but at the same time suggests that it could improve further. IKEA’s emphasis on continuous improvement is mirrored in the very title of the report ‘The never ending job’, which underscores the company’s constant effort towards sustainability.

One major difference between the two reports concerns the frequency of the instances of Affect. The data on the distribution of the categories of Attitude (Table 4) shows that Affect terms are significantly more frequent in IKEA’s text. This difference is also reflected in the number of instances of authorial Affect reported in Table 5. Whereas BP’s list features only one, confidence, IKEA’s list includes four: want, ambition, hate, concern. It is noteworthy to point out that not all of them are positive. The latter two, hate and concern, are instances of negative Affect and are used in association with negative or worrying matters, as in:

(33) IKEA is concerned about climate change and wants to reduce carbon dioxide emissions from all aspects of our operations.

The frequent use of Affect terms can be read as an attempt by IKEA to establish empathy and rapport with the reader. By expressing its desires, ambitions and concerns, the company seeks common ground with him or her and narrows the perceived inter-subjective gap that separates them. A striking example of this is the use of the affect verb hate followed by an exclamation mark in the following passage:

(34) At IKEA, we hate waste! This is why we take every opportunity to turn spill from production into raw material for other products.

The analysis of Engagement reveals key differences between the two reports. In general, Engagement markers are comparatively more frequent in IKEA’s text (Table 7), although this difference is not statistically significant.

Table 7.

Engagement markers^a

	Proclaim	Entertain
BP	4.07	5.29
IKEA	4.62	6.26

Occurrences per thousand words; χ² = 0.021, df = l, p = 0.884.

Significant differences are evident in the distribution of Engagement markers into word classes. Table 8 shows that verbal markers of Proclaim, predictive, possibility and necessity modals are more frequent in IKEA’s text, while BP’s text features a higher number of verbal markers of Entertain.

Table 8.

Markers of Engagement: Word classes^a

		BP	IKEA
Proclaim*	adverbs	0.25	-
	verbs	0.62	1.22
	adjectives	0.08	-
	predictive modals	3.12	3.40
Entertain*	verbs	2.01	1.25
	adjectives	0.08	0.03
	possibility modals	2.75	2.87
	necessity modals	0.45	2.11

Occurrences per thousand words; * p < 0.001, Fisher’s exact test.

The verbs show and see rank highest in IKEA’s list of verbal markers of Proclaim (Table 9). While the frequency of show is just slightly higher in IKEA’s text, see is never used as evidential marker of Proclaim in BP’s report. Direct sensory evidence, as noted by Chafe (1986: 267), signals in English high reliability of knowledge, and therefore contracts the dialogic space for discursive interaction and confrontation. In IKEA’s text this verb is consistently employed to introduce positive evaluative statements regarding progress among suppliers, as, for example, in:

(35) Since IKEA introduced its code of conduct IWAY in 2000, we have seen continuous improvements at IKEA suppliers.

Table 9.

Markers of Proclaim

	BP	Fr.^a	IKEA	Fr.^a
Adverbs	certainly	0.08	-
	in fact	0.08
	no doubt	0.04
	indeed	0.04
Verbs	show	0.37	show	0.56
	demonstrate	0.21	see	0.53
	prove	0.04	prove	0.10
			know	0.03
Adjectives	evident	0.04	-
	true	0.04
Predictive modals	will	2.63	will	2.94
	would	0.45	would	0.33
	shall	0.04	shall	0.13

Occurrences per thousand words.

Along with the frequent use of the verb see to report information on suppliers’ conduct is the frequent employment of necessity modals, in particular must, to refer to the obligations with which they are expected to comply (Table 10). Consider, for example:

(36) IKEA food suppliers must comply with the requirements in our code of conduct IWAY, but also with specific industry-related requirements specified in a supplement to IWAY.

Table 10.

Markers of Entertain

	BP	Fr.^a	IKEA	Fr.^a
Verbs	believe	1.07	expect	0.49
	estimate	0.29	believe	0.43
	expert	0.25	see	0.23
	indicate	0.16	assume	0.03
	see	0.16	indicate	0.03
	assume	0.04	estimate	0.03
	think	0.04
Adjectives	alleged	0.04	likely	0.03
	likely	0.04
Possibility modals	can	1.89	can	2.38
	could	0.53	could	0.33
	may	0.29	may	0.17
	might	0.04
Necessity modals	should	0.29	must	1.78
	must	0.16	should	0.33

Occurrences per thousand words.

Together, these two markers construe IKEA’s relationship with its suppliers as one of power and control. IKEA is presented as an authoritative and attentive supervisor, which provides suppliers with clear guidelines for responsible conduct and objectively reports on their performance. Suppliers must follow IKEA’s normative guidelines and are evaluated against them. Assessments are based on direct, sensory evidence, which lends them credibility and objectivity.

The frequent use of see and must not only defines the relationship between IKEA and its suppliers, but is also central to the discursive management of IKEA’s responsibility towards its stakeholders. Even if dialogically contracting, the verb see still frames IKEA’s assessments of suppliers’ conduct as contingent and subjective. Their intensity is lower than fully endorsed bare utterances. Compare, for instance, the sentence in example 35 earlier with:

(37) * Since IKEA introduced its code of conduct IWAY in 2000, there have been continuous improvements at IKEA suppliers.

The employment of the evidential see hedges IKEA’s responsibility for its suppliers’ behaviour, implicitly acknowledging the possibility that its controls fail and suppliers override its precepts. The frequent use of must can be explained in a similar way. While IKEA can provide a detailed account of the standards its suppliers are expected to meet, it cannot faithfully declare that they fully comply with its requirements. The deontic must performs in this sense a hedging function, underlying IKEA’s resolution but leaving options open for the possibility that suppliers may fail to conform to the company’s norms.

Another clear difference between the two reports resides in the frequency and use of predictive modals. The quantitative data on the markers of Proclaim show that predictive modals are comparatively more frequent in IKEA’s text. This difference, however, is relatively minor. A more substantial difference emerges from the qualitative analysis of the content of predictions. Inspection of the KWIC list for will in the two texts reveals that, while IKEA’s predictions mostly relate to the impact of its decisions on the achievement of its goals, as in:

(38) We are developing a number of tools that will help us to make it easier for customers to leave their cars at home.

(39) Some 10 million children will benefit from IKEA Social Initiative supported programmes,

BP makes instead frequent ‘futurological’ predictions, the scope of which goes well beyond the companies’ own sphere of action to include suppositions about the future of energy supplies, future trends in the energy market, technological development and sustainability. Consider, for example:

(40) Fossil fuels will continue to play a major part in meeting the energy challenges, providing a continuing role for BP’s core business.

(41) We estimate that demand for energy will rise significantly in the future.

Predictions like these are of striking contemporary relevance, address powerful interest groups, among whom are government regulators and financial stakeholders, and have crucial implications for society as a whole. Given their generality and the responsibility they imply, the very fact that these predictions are made is highly significant. Fairclough (2003) underlines the connection between futurological predictions and social power. According to the author,

the power of making futurological predictions is a significant one, because injunctions about what people must do or must not do now can be legitimized in terms of such predictions about the future, and extensively are. (Fairclough, 2003: 167)

Certain claims about the future can be used to justify, legitimate or persuade for certain courses of action in the present. Clearly, to be effective in their legitimating function, futurological predictions need to be grounded in the authority, real or perceived, of the person or institution making such predictions. The consistent use of Appraisal resources to foreground BP’s expertise can be seen as functional to the construction of the basis of perceived authority that is necessary to advance credible, and thus effective, legitimating predictions. At the same time, the very exercise of the power of futurological prediction contributes to the establishment of BP’s expert authority (Fairclough, 2003).

Another substantial difference between the two reports concerns the dialogic expanding verb believe. This verb is strikingly more frequent in BP’s text, where it is persistently used to frame the company’s viewpoint on hotly debated issues which are likely to significantly affect its interests but the scope and implications of which, as in the case of predictions discussed above, exceed BP’s responsibilities and sphere of action. Consider, for example:

(42) Climate change is a major global issue, one which justifies precautionary action in pursuit of a long-term goal along with a programme of action to deliver it. BP believes both governments and industry need to play their parts in achieving such a goal: governments by setting an appropriate policy framework and companies by investing within that framework to deliver a sustainable energy mix.

In this excerpt BP overtly calls upon governments to take action and cooperate with companies to tackle the problem of climate change. The responsibility implied in such a statement and the implications in terms of the power relations between the private and public sector explain the use of a dialogic expanding marker of Engagement. Yet, while believe opens the dialogic space to alternative viewpoints, framing BP’s statement as a contingent and subjective proposal, it also presents it as grounded in BP’s own opinion, whose reliability has been established throughout the text via the repeated use of Appraisal to underscore the company’s expert knowledge. In other words, similarly to predictions, BP’s use of Appraisal resources contributes to support the credibility of its statements. The resulting rhetorical effect is that of ‘soft’ persuasion, whereby the reader is pulled towards the company’s truth by virtue of its trustworthiness and expert insight. By foregrounding its technical expertise, BP establishes its authority on the issues discussed and promotes the acceptance of its viewpoint.

7. Conclusion

The analysis has revealed two radically different approaches to the discursive construction of a responsible corporate identity and two markedly different ways of approaching the reports’ readerships. While BP deploys interpersonal resources to portray itself as a reliable and authoritative expert, IKEA discloses itself as a sensitive and caring corporation, engaged in a continual effort to improve. Further, while BP establishes in discourse a detached, pragmatic relationship with its interlocutors, characterized by a lack of emotional display and an emphasis on technical excellence that conceals the company’s ‘humanity’, IKEA empathizes with the reader and seeks to bond with him or her, establishing a common ground of shared values and feelings. What remains to be discussed is how BP’s and IKEA’s use of Appraisal can be related to their communicative goals. In other words, assuming that the BP and IKEA reports are motivated by a desire to be perceived as legitimate by their stakeholders, how does the particular relational identity they put forward aid their legitimization effort? In this last section I aim to draw some general conclusions and provide a tentative answer to this question.

Suchman (1995) identifies and describes several legitimacy-maintenance strategies. One of these strategies consists in the communication of particular personal attributes aimed at eliciting a positive attitudinal response from receivers. According to the author, in order to maintain their legitimacy, ‘organizations may stock-pile goodwill and support. Generally, such stockpiles are dispositional in character, reflecting either pragmatic attributions (such as trust) or moral attributions (such as esteem)’ (Suchman, 1995: 596). BP’s and IKEA’s use of interpersonal resources can be seen as compatible with these two options respectively. BP’s emphatically authoritative attitude can be interpreted as an attempt to elicit trust from its interlocutors based on the company’s proclaimed competence, in plain words: ‘you can trust me because I know what I am doing’. On the other hand, IKEA’s reader-oriented accommodating stance, ‘I understand your needs and I strive to constantly improve’, can be read as aiming for receivers’ respect and esteem.

But what effects might be sought through these strategies? What consequences could they have for the companies’ legitimacy? In order to answer this question, we first need to consider who BP and IKEA are mainly talking to. According to legitimacy theory, legitimacy is a resource that companies need in order to continue to operate and exist. Legitimacy is conferred to organizations by society upon compliance with its norms and expectations. However, several authors have suggested that certain stakeholder groups within society are more influential than others in determining an organization’s legitimacy (e.g. Neu et al., 1998; O’Donovan, 2002). O’Donovan (2002: 347) maintains that, in order to successfully manage their legitimacy, companies need to identify relevant issues or events and, at the same time, identify those groups of stakeholders ‘who have the necessary attributes to be able to confer or withdraw legitimacy on the corporation in respect of those issues/events’. As Neu et al. (1998) suggests, companies will primarily address those stakeholders that are most relevant to their legitimacy and devote the greatest communicative effort to manage their relationship with them. Although their focus is on the amount of information disclosed, Neu et al. (1998) provide empirical evidence for the assumption that the relative power of different publics encourages differential disclosure responses. Considering who BP’s and IKEA’s relevant publics are is thus important to understand their communicative strategies.

In the case of BP, I propose that financial stakeholders and regulators are the most relevant and potentially influential groups and thus the company’s primary interlocutors. BP works in a capital intensive industry and thus crucially relies on private investments. On the other hand, the company also relies on government concessions to operate, and the magnitude of its environmental impact is such that its activity is under constant government scrutiny. Moreover, the energy sector is heavily affected by the decisions made by policy-makers, who need to respond to public concerns and, at the same time, regulate the market to ensure a reliable and affordable supply of energy. The company’s communicative efforts will thus be primarily directed at negotiating the competing demands from investors on the one hand and regulators on the other. In the case of IKEA, I suggest that customers are the company’s most relevant public. This assumption is based on the simple consideration that if they stop buying IKEA’s products, the company’s profits will suffer. Therefore, the company will primarily address its customers’ concerns and attempt to pre-empt the impact of potential adverse reports on their behaviour.

In light of these considerations, the following conclusions may be drawn. By persistently emphasizing its technical knowledge and expertise, making futurological forecasts and giving insider opinions, BP construes its role vis-a-vis regulators as ‘the expert’, qualifying its relationship with them as one of distance and separation of competencies. This, in turn, can aid the company’s legitimacy management efforts in several ways. By adopting an authoritative stance, BP can foster regulators’ trust. Based on this trust, the company can forestall increased regulation and controls, gaining more autonomy of action. Further, by adopting an authoritative stance, BP can gain more credibility and thus more influence on policy-making processes. More autonomy and more influence imply that the company is better able to pursue its economic interests, protect its investments and thus guarantee future profits for investors. This, in turn, can help the company secure financial support. Public concerns are also addressed in this way. By adopting an authoritative stance, BP is able to buffer the impact of the pressures on regulators coming from other stakeholders, for example, environmentalists and public opinion, by setting its own expert – and thus more credible – opinion against theirs.

On the other hand, by emphasizing its focus on progress, its tenacity in the pursuit of desirable social goals and overtly expressing its desires and concerns, IKEA construes its relationship with its customers as one of empathy, proximity and alignment. The company embraces their concerns, implicitly recognizing them as legitimate, and responds by communicating its adherence to the values of commitment, progress and tenacity. The implications for the company’s legitimacy are twofold. First, by empathizing with its interlocutors and communicating its commitment towards shared goals, IKEA can enhance customers’ goodwill and support. Customers perceive that their worries are not dismissed or ignored, but understood and shared by IKEA. Second, by stressing progress instead of achievements, IKEA implicitly admits the possibility of failure. This can have the effect of hedging the company’s responsibility for future negative events and cushioning their impact on its reputation and legitimacy.

In conclusion, through Appraisal resources, BP and IKEA mould their corporate identity and negotiate their relationship with their most relevant publics. The companies discursively construe their identity and relationship with their legitimacy-conferring stakeholders in a way that, given their specific environment and the relationship they entertain with them, can best enhance their legitimacy. Indeed, through Appraisal resources BP and IKEA do not just modulate their legitimating response, but also discursively negotiate their social responsibility.

Footnotes

Acknowledgements

Earlier versions of this article were presented at the first ‘Discourse Ideology and Society’ Young Scholars Symposium in Łódź, Poland and at the ICAME 32 Conference, Oslo, Norway. I would like to thank Marco Baroni for his continued support and helpful criticism throughout the preparation of this article. I would also like to thank Catherine Riley, for her copyediting work. Special thanks also go to Marina Bondi, Ericka Costa and to the reliability coders Chiara Conter and Anna Feltracco.

Notes

References

Adams

Harte

(1998) The changing portrayal of the employment of women in British banks’ and retail companies’ corporate annual reports. Accounting, Organizations and Society 23(8): 781–812.

Arnold

(1990) The state and political theory in corporate social disclosure research: A response to Guthrie and Parker. Advances in Public Interest Accounting 3(2): 177–181.

Artstein

Poesio

(2008) Inter-coder agreement for computational linguistics. Computational Linguistics 34(4): 555–596.

Bakhtin

(1981) The Dialogical Imagination. Austin: University of Texas Press.

Bebbington

Larrinaga

Moneva

(2008) Corporate social reporting and reputation risk management. Accounting, Auditing & Accountability Journal 21(3): 337–361.

Bednarek

(2006) Evaluation in Media Discourse: Analysis of a Newspaper Corpus. London: Continuum International Publishing Group Ltd.

Biber

Finegan

(1988) Adverbial stance types in English. Discourse Processes 11(1): 1–34.

Biber

Finegan

(1989) Styles of stance in English: Lexical and grammatical marking of evidentiality and affect. Text: Interdisciplinary Journal for the Study of Discourse 9(1): 93–124.

Biber

Johansson

Leech

Conrad

Finegan

Quirk

(1999) Longman Grammar of Spoken and Written English. Cambridge, MA: MIT Press.

10.

Brown

Deegan

(1998) The public disclosure of environmental performance information: A dual test of media agenda setting theory and legitimacy theory. Accounting and Business Research 29(1): 21–41.

11.

Buhr

(1998) Environmental performance, legislation and annual report disclosure: The case of acid rain and Falconbridge. Accounting, Auditing & Accountability Journal 11(2): 163–190.

12.

Buhr

Reiter

(2006) Ideology, the environment and one world view: A discourse analysis of Noranda’s environmental and sustainable development reports. In: Freedman

Jaggi

(eds) Advances in Environmental Accounting & Management, vol. 3. Bradford: Emerald Group Publishing Limited, 1–48.

13.

Campbell

Craven

Shrives

(2003) Voluntary social reporting in three FTSE sectors: A comment on perception and legitimacy. Accounting, Auditing & Accountability Journal 16(4): 558–581.

14.

Carroll

Shabana

(2010) The business case for corporate social responsibility: A review of concepts, research and practice. International Journal of Management Reviews 12(1): 85–105.

15.

Chafe

(1986) Evidentiality in English conversation and academic writing. In: Chafe

Nichols

(eds) Evidentiality: The Linguistic Coding of Epistemology. Norwood, NJ: Ablex Publishing Corporation, 261–272.

16.

Chafe

Nichols

(eds) (1986) Evidentiality: The Linguistic Coding of Epistemology. Norwood, NJ: Ablex Publishing Corporation.

17.

Cohen

(1960) A coefficient of agreement for nominal scales. Educational and Psychological Measurement 20(1): 37–46.

18.

Conrad

Biber

(2000) Adverbial marking of stance in speech and writing. In: Hunston

Thompson

(eds) Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: Oxford University Press, 56–73.

19.

Deegan

(2002) Introduction: The legitimising effect of social and environmental disclosures – a theoretical foundation. Accounting, Auditing & Accountability Journal 15(3): 282–311.

20.

Deegan

Gordon

(1996) A study of the environmental disclosure practices of Australian corporations. Accounting and Business Research 26: 187–199.

21.

Deegan

Rankin

(1996) Do Australian companies report environmental news objectively? An analysis of environmental disclosures by firms prosecuted successfully by the environmental protection authority. Accounting, Auditing & Accountability Journal 9(2): 50–67.

22.

Deegan

Rankin

Tobin

(2002) An examination of the corporate social and environmental disclosures of BHP from 1983–1997: A test of legitimacy theory. Accounting, Auditing & Accountability Journal 15(3): 312–343.

23.

Deegan

Rankin

Voght

(2000) Firms’ disclosure reactions to major social incidents: Australian evidence. Accounting Forum 24(1): 101–130.

24.

Dowling

Pfeffer

(1975) Organizational legitimacy: Social values and organizational behavior. Pacific Sociological Review 18(1): 122–136.

25.

Eggins

(2004) An Introduction to Systemic Functional Linguistics. London: Continuum International Publishing Group.

26.

Eggins

Slade

(2005) Analysing Casual Conversation. London: Equinox Publishing Ltd.

27.

Fairclough

(1992) Discourse and Social Change. Cambridge: Polity Press.

28.

Fairclough

(2003) Analysing Discourse: Textual Analysis for Social Research. London: Routledge.

29.

Gray

(2002) The social accounting project and Accounting Organizations and Society: Privileging engagement, imaginings, new accountings and pragmatism over critique? Accounting, Organizations and Society 27(7): 687–708.

30.

Gray

Javad

Power

Sinclair

(2001) Social and environmental disclosure and corporate characteristics: A research note and extension. Journal of Business Finance & Accounting 28(3–4): 327–356.

31.

Gray

Kouhy

Lavers

(1995) Corporate social and environmental reporting: A review of the literature and a longitudinal study of UK disclosure. Accounting, Auditing & Accountability Journal 8(2): 47–77.

32.

Gray

Owen

Adams

(1996) Accounting and Accountability: Changes and Challenges in Corporate Social and Environmental Reporting. London: Prentice Hall.

33.

Halliday

(1978) Language as Social Semiotic: The Social Interpretation of Language and Meaning. London: Edward Arnold.

34.

Halliday

(1994) An Introduction to Functional Grammar. London: Edward Arnold.

35.

Hood

(2004) Appraising research: Taking a stance in academic writing. PhD thesis, Faculty of Education, University of Technology, Sydney. Available at: http://www.grammatics.com/appraisal/hoodS-phd-links.htm (accessed 18 May 2011).

36.

Hoye

(1997) Adverbs and Modality in English. London: Longman.

37.

Hunston

(1993) Evaluation and ideology in scientific writing. In: Ghadessy

(ed.) Register Analysis: Theory and Practice. London: Pinter Publishers, 57–73.

38.

Hunston

(1994) Evaluation and organization in a sample of written academic discourse. In: Coulthard

(ed.) Advances in Written Text Analysis. London: Routledge, 191–218.

39.

Hunston

(2004) Counting the uncountable: Problems of identifying evaluation in a text and in a corpus. In: Partington

Morley

Haarman

(eds) Corpora and Discourse. New York: Peter Lang, 157–188.

40.

Hunston

Thompson

(2000) Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: Oxford University Press.

41.

Hyland

(1996) Writing without conviction? Hedging in science research articles. Applied Linguistics 17(4): 433–454.

42.

KPMG (2008) International Survey of Corporate Sustainability Reporting. Available at: http://www.kpmg.com/ (accessed 14 January 2011).

43.

Krippendorff

(2004) Content Analysis: An Introduction to its Methodology. Thousand Oaks, CA: SAGE.

44.

Laine

(2005) Meanings of the term ‘sustainable development’ in Finnish corporate disclosures. Accounting Forum 29(4): 395–413.

45.

Laine

(2010) Towards sustaining the status quo: Business talk of sustainability in Finnish corporate disclosures 1987–2005. European Accounting Review 19(2): 247–274.

46.

Landis

Koch

(1977) The measurement of observer agreement for categorical data. Biometrics 33(1): 159–174.

47.

Lischinsky

(2010) The struggle over sustainability: A corpus approach to managerial conceptions of sustainable development. In: Proceedings of the Critical Approaches to Discourse Analysis Across the Disciplines Conference, 13–15 September, Łódź, Poland.

48.

Livesey

(2002) The discourse of the middle ground: Citizen Shell commits to sustainable development. Management Communication Quarterly 15(3): 313–349.

49.

Livesey

Kearins

(2002) Transparent and caring corporations? A study of sustainability reports by the Body Shop and Royal Dutch/Shell. Organization & Environment 15(3): 233–258.

50.

Martin

(1995) Reading positions/positioning readers: Judgment in English. Prospect 10(2): 27–37.

51.

Martin

(2000) Beyond exchange: APPRAISAL systems in English. In: Hunston

Thompson

(eds) Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: Oxford University Press, 142–176.

52.

Martin

White

(2005) The Language of Evaluation: Appraisal in English. London and New York: Palgrave Macmillan.

53.

Milne

Kearins

Walton

(2005) Playing with magic lanterns: The New Zealand Business Council for Sustainable Development and corporate triple bottom line reporting. In: Proceedings of 4 APIRA Conference, Singapore, 4–6 July.

54.

Milne

Tregidga

Walton

(2009) Words not actions! The ideological role of sustainable development reporting. Accounting, Auditing & Accountability Journal 22(8): 1211–1257.

55.

Neu

Warsame

Pedwell

(1998) Managing public impressions: Environmental disclosures in annual reports. Accounting, Organizations and Society 23(3): 265–282.

56.

O’Donovan

(2002) Environmental disclosures in the annual report: Extending the applicability and predictive power of legitimacy theory. Accounting, Auditing & Accountability Journal 15(3): 344–371.

57.

Page

(2003) An analysis of appraisal in childbirth narratives with special consideration of gender and storytelling style. Text: Interdisciplinary Journal for the Study of Discourse 23(2): 211–237.

58.

Palmer

(2001) Mood and Modality. Cambridge: Cambridge University Press.

59.

Patten

(1992) Intra-industry environmental disclosures in response to the Alaskan oil spill: A note on legitimacy theory. Accounting, Organizations and Society 17(5): 471–475.

60.

Roberts

(1992) Determinants of corporate social responsibility disclosure: An application of stakeholder theory. Accounting, Organizations and Society 17(6): 595–612.

61.

Shocker

Sethi

(1973) An approach to incorporating societal preferences in developing corporate social action. California Management Review 15(4): 97–105.

62.

Skulstad

(2008) Creating a green image in the public sphere: Corporate environmental reports in a genre perspective. In: Wodak

Koller

(eds) Handbook of Communication in the Public Sphere. Berlin: De Gruyter Mouton, 181–201.

63.

Suchman

(1995) Managing legitimacy: Strategic and institutional approaches. The Academy of Management Review 20(3): 571–610.

64.

Thomson

(2007) Mapping the terrain of sustainability accounting. In: Unerman

Bebbington

O’Dwyer

(eds) Sustainability Accounting and Accountability. New York: Routledge, 19–37.

65.

Tilling

Tilt

(2010) The edge of legitimacy: Voluntary social and environmental reporting in Rothmans’ 1956–1999 annual reports. Accounting, Auditing & Accountability Journal 23(1): 55–81.

66.

Tinker

Neimark

Lehman

(1991) Falling down the hole in the middle of the road: Political quietism in corporate social reporting. Accounting, Auditing & Accountability Journal 4(2): 28–54.

67.

Ullmann

(1985) Data in search of a theory: A critical examination of the relationships among social performance, social disclosure, and economic performance of US firms. The Academy of Management Review 10(3): 540–557.

68.

White

(2001) An introductory tour through Appraisal theory (word processor version). Available at: http://www.grammatics.com/appraisal (accessed 18 May 2011).

69.

White

(2003) Beyond modality and hedging: A dialogic view of the language of intersubjective stance. Text: Interdisciplinary Journal for the Study of Discourse 23(2): 259–284.